Do imputed education histories provide satisfactory results in fertility analysis in the Western German context?



Demographic Research: Volume 21, Article 6

be more frequently misallocated to the ‘in university education’ category for the
childless. This would again bias the estimate for ‘vocational degree’ upward, since
those who are childless would be selectively taken out of this category in the imputed
histories. This happens to only a very small extent in the current imputations, as can be
seen for instance in Table A1a in the appendix. Only 1% of the exposure time originally
spent holding a vocational degree is imputed as time spent enrolled in university
education. This is because, as already mentioned above, in the cohort analyzed here, it
is very uncommon to begin university education after having already obtained a
vocational degree. This is not very likely to change in the future, since an upper
secondary school degree is generally required to enroll at universities, which is an
obstacle for many people. In other countries though, it may be more common to take up
university education later in the life course after already having held a lower level
degree for some time. In those contexts, imputations would be more problematic.

A similar problem might turn up in Germany, however, if more women obtain a
master craftswoman’s or technician’s degree after their basic vocational degree. In the
cohort analyzed here, only 8% had a master craftswoman’s or technician’s degree. If in
later cohorts, it becomes more common for women to obtain these higher level
vocational degrees, and if they often do so later in the life course, imputation might
become more problematic for reasons parallel to those described above for obtaining
university degrees after vocational degrees. In the models estimated in this study, those
with master craftswoman’s or technician’s degrees were not distinguished from those
with basic vocational degrees. It might nonetheless make a difference for the estimates
whether this differentiation is explicitly made in the questionnaire or not. If people are
asked to give the first date they obtained their highest degree, and master craftswoman’s
degrees are explicitly listed in the questionnaire, those who have a master
craftswoman’s degree will choose that category (instead of just ‘vocational degree’) as
their highest degree. Even if they are asked to give the first date they obtained their
highest degree, they will give the date they obtained their master craftswomen’s degree
instead of the date they obtained their basic vocational degree. The third and the fourth
imputations simulate that case, assuming the questionnaire included differentiated
degree categories (Table 2). Here we can see that in the third imputation, that assumes
the questionnaire asked for the first date the respondent received her highest degree but
provides differentiated degree categories, estimates for the baseline are between those
for the first and second imputation. Using the first date the highest degree was obtained
works towards underestimating the baseline like in the first imputation. But, using the
date respondents obtained a master craftswomen’s degree (for respondents who have
that degree) instead of the date they first obtained their basic vocational degree, works
toward overestimating the baseline. Since only few respondents in this sample actually
had a master craftswoman’s degree, it appears that the result was altogether still a slight

http://www.demographic-research.org

153




More intriguing information

1. Investment in Next Generation Networks and the Role of Regulation: A Real Option Approach
2. The name is absent
3. EU enlargement and environmental policy
4. WP 36 - Women's Preferences or Delineated Policies? The development or part-time work in the Netherlands, Germany and the United Kingdom
5. The name is absent
6. The Impact of Hosting a Major Sport Event on the South African Economy
7. The name is absent
8. Subduing High Inflation in Romania. How to Better Monetary and Exchange Rate Mechanisms?
9. The name is absent
10. The name is absent
11. Knowledge, Innovation and Agglomeration - regionalized multiple indicators and evidence from Brazil
12. How Offshoring Can Affect the Industries’ Skill Composition
13. The name is absent
14. The Challenge of Urban Regeneration in Deprived European Neighbourhoods - a Partnership Approach
15. The name is absent
16. Giant intra-abdominal hydatid cysts with multivisceral locations
17. The storage and use of newborn babies’ blood spot cards: a public consultation
18. The name is absent
19. Consumer Networks and Firm Reputation: A First Experimental Investigation
20. GENE EXPRESSION AND ITS DISCONTENTS Developmental disorders as dysfunctions of epigenetic cognition